Starting Pitcher Game Ratings

Back to Home Page

Back to Members Index

Our Game Ratings are designed to supplement the won-loss and ERA statistics, as well as the quality starts. Game Ratings range from 0 to 5 points and are recorded for each game using these rules:

a) one point for pitching 6 innings, no points at all for the game if the starter does not go 5 innings. Starters going 5.00, 5.33 or 5.67 innings can get points but cannot get more than 4, as they get no point for innings.

b) one point for hits equal to or fewer than innings.

c) one point for strikeouts equalling innings minus 2 or better. Six innings needs 4 strikeouts for a point, 6.33 needs 5 strikeouts.

d) one point for strikeouts double or more the number of walks.

e) one point for yielding zero or only one home runs.

This method has the benefit of being simple and easy to apply using ordinary newspaper boxscores. By taking into account strikeouts, walks and home runs, it utilizes the pitching results that are not dependent on fielding, earned runs, or on bullpen support, common criticisms of ERA and quality starts.

When large numbers of games are considered, the Game Ratings follow a pattern that is very consistent over time.

This same pattern in ERA and winning percentage has appeared every time we did this exercise, no matter the league, or seasons, or even part-season. In 2008, the winning percentage for GR 5 games is down a little - perhaps only temporarily, through games of May 26th..

Sometimes the individual GRs can be a little jarring. A pitcher who gives up several runs may still get a good GR, if his underlying pitching performance is good. This season two pitchers have given up 5 earned runs in a start and still gotten a GR of 5 for that game, while two others have yielded no runs in 6 or more innings but only pitched well enough for a 2 Game Rating. Do not be thrown of course by such individual aberrant games. The Game Ratings are almost as good as detailed sophisticated methods that you may have heard about like Component ERA and Expected ERA. And all of them are better at predicting future ERA than ERA is as a predictor of future ERA!

Pitchers who throw many starts that are 4s and 5s are dominators, while disaster starts are the 0s and 1s. Generally speaking, 50% or better dominating starts is good, and under 18% disasters is good. The better pitchers will hit or better those marks.The Cy Young winners last year were 71%-6% and 67%-0% Erratic pitchers will have more than 18% disasters but also 50% or more dominating games. Dependable pitchers have less than 50% dominating, but have few disasters. Pitchers under 50% dominating and over 18% disasters are suspect, and those with more disasters than 4s/5s are quite suspect - and may be in danger of losing their jobs.

A regular rotation starter who throws two consecutive 4/5 starts is likely to continue pitching well. In a survey of over 400 such starts, following two 4/5 games, pitchers had a higher W-L percentage and a better ERA. Other surveys have noted that even one 4/5 start is fairly likely to be followed by another or at least a start good enough for a team win..

The very best pitchers were even more likely to continue than the average starters. And they were very good bets in road games, often undervalued by the betting line - and were superb as away underdogs. In fact, characterizing starters based on their prior year records (more than 50% dominating, under 17% disasters listed individually in the Baseball Insight Annual) produced very good results v. the line on the road.

A regular starter, not the very best, who pitches a 0/1 disaster game is likely to continue below par for at least one more game, although not necessarily a disaster. Such pitchers were much more apt to ‘bounce back’ after a disaster start when they were at home, and they had a good record as home underdogs, at least in a survey of over 200 such games.

So one way of using the Game Ratings is to look for consecutive 4/5 games, and watch out for those 0/1 starts. In the weekly we have room for only the last ten starts by each listed pitcher. An easy shorthand method might be to look for a pitcher with five games of 4/5 in his last ten starts and no or only one 0/1 game in that span. Such hurlers are pitching well. The last six gives a good handle on current form - look for three 4/5 games in those, no or only one 0/1 game.